Files
openclaw/src/plugins/web-content-extractors.runtime.ts
2026-04-26 03:47:45 +01:00

125 lines
4.2 KiB
TypeScript

import type { OpenClawConfig } from "../config/types.openclaw.js";
import { resolveBundledPluginCompatibleLoadValues } from "./activation-context.js";
import {
createPluginActivationSource,
normalizePluginsConfig,
resolveEffectivePluginActivationState,
} from "./config-state.js";
import type { PluginManifestRecord } from "./manifest-registry.js";
import { loadPluginManifestRegistryForPluginRegistry } from "./plugin-registry.js";
import { loadBundledWebContentExtractorEntriesFromDir } from "./web-content-extractor-public-artifacts.js";
import type { PluginWebContentExtractorEntry } from "./web-content-extractor-types.js";
function compareExtractors(
left: PluginWebContentExtractorEntry,
right: PluginWebContentExtractorEntry,
): number {
const leftOrder = left.autoDetectOrder ?? Number.MAX_SAFE_INTEGER;
const rightOrder = right.autoDetectOrder ?? Number.MAX_SAFE_INTEGER;
if (leftOrder !== rightOrder) {
return leftOrder - rightOrder;
}
return left.id.localeCompare(right.id) || left.pluginId.localeCompare(right.pluginId);
}
function resolveBundledWebContentExtractorCompatPluginIds(params: {
config?: OpenClawConfig;
workspaceDir?: string;
env?: NodeJS.ProcessEnv;
onlyPluginIds?: readonly string[];
}): string[] {
const onlyPluginIdSet =
params.onlyPluginIds && params.onlyPluginIds.length > 0 ? new Set(params.onlyPluginIds) : null;
return loadPluginManifestRegistryForPluginRegistry({
config: params.config,
workspaceDir: params.workspaceDir,
env: params.env,
includeDisabled: true,
})
.plugins.filter(
(plugin) =>
plugin.origin === "bundled" &&
(!onlyPluginIdSet || onlyPluginIdSet.has(plugin.id)) &&
(plugin.contracts?.webContentExtractors?.length ?? 0) > 0,
)
.map((plugin) => plugin.id)
.toSorted((left, right) => left.localeCompare(right));
}
function resolveEnabledBundledExtractorPlugins(params: {
config?: OpenClawConfig;
workspaceDir?: string;
env?: NodeJS.ProcessEnv;
onlyPluginIds?: readonly string[];
}): PluginManifestRecord[] {
if (params.config?.plugins?.enabled === false) {
return [];
}
const activation = resolveBundledPluginCompatibleLoadValues({
rawConfig: params.config,
env: params.env,
workspaceDir: params.workspaceDir,
onlyPluginIds: params.onlyPluginIds,
applyAutoEnable: true,
compatMode: {
allowlist: true,
enablement: "always",
vitest: true,
},
resolveCompatPluginIds: resolveBundledWebContentExtractorCompatPluginIds,
});
const normalizedPlugins = normalizePluginsConfig(activation.config?.plugins);
const activationSource = createPluginActivationSource({
config: activation.activationSourceConfig,
});
const onlyPluginIdSet =
params.onlyPluginIds && params.onlyPluginIds.length > 0 ? new Set(params.onlyPluginIds) : null;
return loadPluginManifestRegistryForPluginRegistry({
config: activation.config,
workspaceDir: params.workspaceDir,
env: params.env,
includeDisabled: true,
}).plugins.filter((plugin) => {
if (
plugin.origin !== "bundled" ||
(onlyPluginIdSet && !onlyPluginIdSet.has(plugin.id)) ||
(plugin.contracts?.webContentExtractors?.length ?? 0) === 0
) {
return false;
}
return resolveEffectivePluginActivationState({
id: plugin.id,
origin: plugin.origin,
config: normalizedPlugins,
rootConfig: activation.config,
enabledByDefault: plugin.enabledByDefault,
activationSource,
}).enabled;
});
}
export function resolvePluginWebContentExtractors(params?: {
config?: OpenClawConfig;
workspaceDir?: string;
env?: NodeJS.ProcessEnv;
onlyPluginIds?: readonly string[];
}): PluginWebContentExtractorEntry[] {
const extractors: PluginWebContentExtractorEntry[] = [];
for (const plugin of resolveEnabledBundledExtractorPlugins({
config: params?.config,
workspaceDir: params?.workspaceDir,
env: params?.env,
onlyPluginIds: params?.onlyPluginIds,
})) {
const loaded = loadBundledWebContentExtractorEntriesFromDir({
dirName: plugin.id,
pluginId: plugin.id,
});
if (loaded) {
extractors.push(...loaded);
}
}
return extractors.toSorted(compareExtractors);
}